Taiwan: Intel CEO Responds to Nvidia's Claims of Outdated Processors in the AI Era

Intel CEO Pat Gelsinger took to the stage at Computex in Taiwan to discuss new products he hopes will help stem the loss of market share

by Sededin Dedovic
Taiwan: Intel CEO Responds to Nvidia's Claims of Outdated Processors in the AI Era
© Alex Wong / Getty Images

Intel's CEO, Pat Gelsinger, took the stage at the Computex trade show in Taiwan to discuss new products that he expects will help halt the loss of market share to competitors, including the AI leader Nvidia. Intel introduced its new Xeon 6 processors for data centers with more efficient cores, enabling operators to reduce the space needed for a given task to one-third of that required for previous generation hardware.

Like its competitors, from Advanced Micro Devices to Qualcomm, Intel touted performance measurements that showed its new silicon to be significantly better than existing options. AMD and Qualcomm executives had previously used Intel processors for laptops and desktops in their Computex keynotes to demonstrate how far ahead they were in certain aspects of technology.

Gelsinger directly responded to Nvidia CEO Jensen Huang's claim that traditional processors like Intel's are falling behind in the era of artificial intelligence. “Contrary to what Jensen would like you to believe, Moore's Law is alive and well,” he said, emphasizing that Intel will play an important role in the proliferation of AI as a leading supplier of PC chips.

Intel firmly adheres to Moore's Law

“I think it's like the internet 25 years ago, it's that big,” Gelsinger said. “We see this as the fuel driving the semiconductor industry towards reaching a value of $1 trillion by the end of the decade”.

Intel's Gaudi systems, which combine its chips into multi-processor bundles tailored for generative AI training, will be offered by partners such as Dell Technologies and Inventec, Gelsinger said. A bundle with eight Intel Gaudi 2 accelerators will be sold for $65,000.

A more powerful bundle with eight Intel Gaudi 3 accelerators will hit the market at $125,000, with the company estimating that both offerings are more affordable than competing products, according to SCMP. Each of these Gaudi 3 clusters consists of 8,192 accelerators, and Intel estimates they offer up to 40% faster training times for AI models compared to an equivalent cluster of Nvidia H100 GPUs.

Intel also stated that Gaudi 3 will be up to twice as fast as Nvidia H100 in performing AI inference tasks, as measured on popular models made by Meta Platforms and Mistral. These advantages might not be sufficient to dethrone Nvidia as the leader in AI data processing centers.

Intel Gaudi© Bloomberg Technology / Youtube channel

“Performance of each individual accelerator is no longer the most important thing,” said Leonard Li, an analyst at neXt Curve. Nvidia's biggest advantage is its cohesive and integrated ecosystem and proprietary technology such as NVLink, which ensures that their computing clusters operate as one.

“The power lies in the ability to create a massive logical accelerator of enormous size”. Santa Clara-based Intel has led the computer industry for decades, but its revenues have declined over the past two years as it has lagged behind competitors.

Gelsinger, who was brought back to the company three years ago to turn its fortunes around, has spent heavily on revitalizing its offerings and building a network of factories he said will reclaim its leadership in chip design and manufacturing.

Although Intel's sales have stopped declining, analysts do not project a quick recovery, and the company is on track to end 2024 with $20 billion less in revenue than it had in 2021. Meanwhile, Nvidia's sales will double, and AMD will grow by more than 10%, according to estimates, as these companies better capitalize on the wave of spending on AI computing hardware.

“This is the most important time of our shared careers,” Gelsinger said, reiterating the importance of Intel working with partners. “We are made for this moment”. Intel's efforts to regain its footing in the competitive semiconductor market are commendable, especially in an era dominated by rapid AI advancements.

Gelsinger's strategic focus on leveraging Intel's historical strengths in innovation and manufacturing capacity is crucial. However, the challenges Intel faces are multifaceted, involving not only technological advancements but also market perception and agility in adapting to new trends.

The introduction of the Gaudi systems and the adherence to Moore's Law indicate a promising direction, but the company must also focus on fostering a robust ecosystem akin to Nvidia's to ensure sustained growth. Partnerships and collaborative innovation will be key in Intel’s journey to reclaim its leadership position in the semiconductor industry.

Intel's focus on AI is strategically sound given the technology’s increasing prevalence across various sectors. AI's applications, from data centers to consumer electronics, represent a vast market. By offering competitive products like the Gaudi systems, Intel aims to capture a significant share of this market.

The Gaudi systems’ performance metrics, which suggest substantial improvements over Nvidia’s H100 GPUs, are encouraging.